Southern African Field Archaeology 6:82-87. 1997.
NOSE-BLEED IN SHAMANS AND ELAND*
F.G. BUTLER
122 High Srreer, Grahamsrown, 6139
*Accepted for publication September 1997
ABSTRACT
As a result of a re-examination of the ethnogr<iphical evidence tor nose-bleeding during trance dancing, the author proposes that it was not an automatic accompaniment of entry into the state of trance (as is widely believed), hut that it was induced by the deliberate insertion hy the dancer of. powerful snuff into his nostrils.
It is further suggested that the technique was also employed hy shamans/hunters to induce nose-blec;:ds in the eland the "totem" animal of the San over much of southem Africa.
THE DANCE OF BLOOD
The most influential single piece of evidence is from the French missionaries, Arbousset and Dumas ( 1846) who discuss a San dance which was so violent that it is "not unusual to st:e someone sink to the ground, exhausted and covered with blood which pours from his nostrils; it is on this account that the dance is called mokomo, or the dance of blood" (ibid. 246-7). There is a clear cause and effect relationship implied between the physical exertion of the dance and the nose-bleed. Unfortunately there is no close-up, in this impona.nt piece of the ethnogmphy, of the commencement of a nose-bleed.
However, in this review it is shown that medical evidence is against any amount of exercise starting a general nose-bleed in humans. Readers will tind it difficult to recall extensive nose-bleeds caused by mere physical exhaustion in screenings of even violent sports or of the Comrades Marathon. The medical literature on the effects of violent exercise simply does not mention nose-bleed. For instance, Torg et a!. ( 1990) deals with nasal fractures among facial injuries common in many body contact spons. Dr..u et a/. (1985), in what must surely be an authoritative work, has only one index entry under nose 'breathing'. Under "effects of dynamic exercise" (ibid. 42) no mention is made of nose-bleed.
Lewis-Williams & Dowson (1994:211) states that "we know from the ethnography that nasal haemorrhage was experienced by trancing shamans". The trance state itself seems responsible for this medically inexplicable physiological symptom. The head-on collision between medical and ethnographic evidence leads to an re-
examination of the latter.
BLEEK AND NOSE-BLEED
In his article on "Ethnographical evidence relating to trancing" and 'shamans' among the northern and southern
Bushmen Lewis-Williams (1992) refers us to four passages in Bleek (1935; 1936). The references all involve that key clue to the interpretation of rock art - nose- bleeding. Only the briefest summaries are possible.
The first account (Bieek 1935:34) is of a battle between rival sorcerers, of whom the better cures the patient by healing an internal wound, which means inhaling or sniffing his opponent from the body of his patient, and then sneezing him out.
There was a real wound where the other sorcerer was. Of that wound the (good) sorcerer wishes to cure him, that it may not stand open.
Th~refore, h~ is wont to rub the man's wound with his nose. We do not see the wound, for the man's skin is over it. But his nose goes into the man's body to the place underneath. His nose sews up the mouth of the man's wound, where the other sorcerer had eaten of the man's flesh
inside.
When he has finished curing him, he takes the blood in which he had sneezed out the other sorcerer, and paints the man who is ill with it, with blood from his nose.
That probing, rubbing nose is presumably being encouraged to bleed, but can mere rubbing cause a nose- bleed? The blood from the nose, however, has curative properties, and the patient is painted with it. In the second account (Bieek 1935: 12):
When a sorcerer is teaching us, wihen his nose bleeds, he sneezes the blood from his nose into his hand, he makes us smell the blood ... etc.
That sounds unequivocal enough. But a sneeze is not something that fits easily into a deliberate "teaching"
milieu. "He sneezes the blood from his nose into his 11and". It is possible that the hand has first induced the sneeze. This is not the involuntary sneeze all experience, it is a sneeze that can be, and is, taught. It suggests that sorcerers (shamans?) may be a class or group that have secrets which are taught to suitable acolytes.
In the third case (Bieek 1935:9), that of a female sorcerer, the narrator admits that the bleeding has been induced. She "snores" the patient, taking something out of her liver and leaves the hut to lie down outside with it. ..
She comes back in order to heal the hole from which she had taken 'the thing' whil~ blood cam~
from her nose (she cut the thing out of her nose when she went outside).
At first perus<~l, the fourth reference (Bleck 1936: 137) does not seem relevant to the trance dance. It was no doubt a profound parable about brothers and sist~rs and rain. "And his younger sister smelt the springhok bush and bled from the nose" -unless the springbok bush is an irritant that induces very violent sneezing.
SNEEZING
This raises the whole question of sneezing. L::wis- Williams (1981 :78) reminds us:
Among the /Xam curing involved the sniffing out of the patient's body of what was harming him Lloyd, perhaps following Campbell (1815:316) translates this as 'snoring'. After the medicine man had sniffed out the evil, he sneezed out the 'harm's things' which wer~ said to resemble arrows, little sticks (a belil!f still held by the Kung, Marshall pers. comm.) or various animals; and as he did this he suffered a nasal haemorrhage. There are indications that this curing operation was accomplished while the medicine man was in tranc~ (Bie~k 1935: I)."
(My emphasis)
In three of the four Bleek citations the blood is sneezed out. There is widespread belief in many cultures tl1at a properly induced sneeze can expel evil from the body. For instance, in Broster and Bourn's (1981) study of Xhosa medicine, the "healing" of Novuka from a charge of witchcraft ends as follows:
The igqirha thumped her three times on the hack between the shoulder blades, and then placed a pinch of snuff on her nostrils so that she sne~zed
for a full minute. After the sneezing she ceased to weep." (Broster & Bourn 1981:79).
Note the use of snuff to induce the violent sneezimg.
h is generally accepted that Xhosa medicine owed much to the San practise (Jolly 1996) and vice versa.
Now in the fourth Bleek nose-bleed citation we ar~
told that the mer~ sniffing of the springbuck bush causes a nose-bleed.
Scientific corroboration of the toxic effects of one of the several plants called springbuck bush on sheep is available. Watt and Breyer-Brandwijk (1962:240) give a detailed account of the experimental feeding of it to four sheep. All die. We are concerned only with some symptoms of the post mortem: "pronounced acute catarrhal enteritis", "pronounced emphysema of the subcutaneous tissues of the intermandibular region and the ventral aspect of the neck"; "pronounced hyperaemia of the mucosa of th~ main bronchi and trachea with haemorrhages". Ifthese arc the results of the plant in the normal digestive process, what might direct contact with the tissues of the respiratory system do?
Unfortunately, tests of the ~ffects on rats of snuff made from Herria pal/ens conducted by the Rhodes
Univ~rsity Pharmacy D~partment have not confirmed the
exp~ctations arous~d hy Watt and Breyer-Brandwijk (1962).
However, Smith ( 1966:422), under "springbokbossie"
lists not only Hertia pollens, hut "vaalbos" Lopholaena cneorifolia), ''kraalbossie" (Aptosimum steingrocveri and Aprosimum leacorrlrizum). He also mentions
"springhokganna", "springhokkaroo", "springbokklaver",
"springbokmelkbos", and "springhok suring".
Although, as in the Bleek reference four, the nose- bleed from a mere smell might take place, a snuff could have been made, not powdered too finely, but containing small, visible splinters whil:h serv~d two functions.
I. When thrust or sniffed up the nose they would help to break the blood vessels.
2. When sneezed out with the nose-bleed, they would provide visible evidence of the "harm's things" -
"the arrows and little sticks" -that had caused the pain, and which had been supposedly sniffed out by the shaman.
Furtlher reference to the second Bleek quotation reveals an inkresting possibility. When the sorcerer (shaman) is teaching his acolytes,
he sneezes his blood from his nose into his hand, he makes us smell the blood ... "
Why must they smell the blood? Perhaps in ord~r that they may detect the aroma of the springbuck bush. So, the Bleek ethnographic record, while not denying the occurrence of nosc-oleeding during trance dance, suggests that it was induced hy recourse to a professional technique which was not .generally known, and had to be taught to suitable candidates.
The induced nose-bleed may strike some as detracting from the religious seriousness of the shaman's actions or presence, but this need not he so. His blood is real enough. and the effects of inhaling the springbuck bush snuff ar,e probably extremely painful.
But, for the shaman himself, the advantages of the
suggested snuff t!!chniques ar\! obvious. Nos!!-bked is a signal of his professional status, and, with the aid of snuff, he can go into trance without the aid of dance, particularly when engaged in his function as heakr. In the first and third examples of nose-bleed quoted from Bleek (1935: 12, 13) there is no suggestion whatever of the healer dancing himself into a state of trance. HI! has nose-bleed on tap, as it were.
BLOOD ON THE FACE
In his detailed :~nalysis of the Linton Rock Paintings, Lewis-Williams (1988:4) describes a ritual which may have survived into the present.
In th!! Linton paintings three figures have blood falling from the nose. In addition a further nine have red I ines radiating fmm the nose across the face. This puzzling feature was recently explained by a very old, probably last survivor of the southern San ... she said the shamans bled from the nose during the dance, and then smeared the blood back across their faces (Jolly 1986, Lewis-Williams 198 6a). Her demonstration of this action showed that lines of blood would have been facial lines as well as blood falling from the nose. There is thus little room for doubt that the figures with blood falling from the nose and those with red linl!s 011 the face all represent San shamans whose potency has boiled and who are in a trance.
The assumption here is that the bleeding is the rl!sult of the shamans' "potency" heing raised to hailing point by the trance dance. But may they not have taken the nose-bleed snuff? A positive answer would accord well with the generally accepted view that a hand-raised-to- the-nose gesture signals a shaman supposedly entering trance. Why that gesture? Once one considers the nose- bleed-inducing c~pacities of the springbuck bush, and the Bleek sources of this information, it is hard not to bdieve that the hand contains snuff. It is worth looking at the first three Bleek sources once .again. There is no mention of dancing and nose-bleeding, only of nose-bleeding and healing -that is, in circumstances where the blood might he sneezed into the hand, and smelt, and examined by the patients, who would see the "arrows" and "hits of wood".
THE DYING OUT OF THE NOSE-BLEED AT TRANCE DANCES
San witnesses (via D Bleek) support the probability that nose-bleeding was not something that tnmce dancers w~re
subject to, but something they induced when th!!y wl!rl!
ready to "di!!", or needed blood for healing purposl!s.
Katz (1982) gives us sevl!ral ·detailed and vivid accounts of dancl!s, and makes extensive use of the personal evidence of trancers (peopk who kia). He devotes twenty pages (ibid: 59-79) to reporting at close
intervals the progress of an all night trance dance. There are some bizarre incid!!nts and episodes involving physical violence which could easily have resulted in bleeding noses. But people trance dance or kia without the benefit of blood. Biesele, (1993 :74-76) gives a general description of trance dances. No one's nose- bkeds.
What has happened? Th!!re is no doubt about the fact that bleeding from the nose was widely witnessed (in fact, practised) at trance dances in the past. Why had it gone out of fashion?
It is possible that, in inducing a nose-bleed, the shamans or trancers were originally simulating the last stages of the familiar running down of an animal. Such running down stopped fairly abruptly in most of South Africa early in the 19th Century, when San society was deliberately destroyed and game became scarce, except in the far Kalahari. and East Griqualand. Dorothea Bleek (1928), in her account of the Nharon of the central Kalahari, mak!!s no ml!ntion of nose-bleeding.
ELAND NOSE-BLEED
It is gl!nerally accepted that the eland is the most commonly depicted animal in Southern African rock art.
During the last third of this century, students of rock art have come to believe that it was not merely hunted for the meat its huge body could supply, but for spiritual power it imparted during its dying. This potency was present particularly in its blood, the material vehicle of its life, and the medium which the shaman/artists employed to mix the paints with which they depicted animals, notably eland; and that such pictures continued to exert spiritual intluence on those who looked at them.
The !hunt of such a quarry was thus an ordinary affair, but had a profoundly "religious" character.
In an attempt to interpret a group of puzzling paintings it is suggested that such hunts (as with all such hunts) were organised by leaders, who were most probably shamans, supported by companions who shared a common ritual, certainly a common dance. It is suggested here that th!!re is evidence that chief huntsman/
shaman sometimes wore a sp!!cial costume, as may some of his colleagues although this is disputed by Lewis- Williams (1981:71). In support of this we have a copy of a painting by Stow and Bleek (1930, tig. 57) identified as a hunting group by an old Bushman informant (Fig. 1).
Before proceeding, however, attention should be drawn to San awareness of the power of the sense of smell, both to attract and to repel. They used a scented h!!rb(s) call!!d buchu to make themselves attractive. They believed that animals had sense equipment similar to their own. This is a key to the understanding of a famous painting in the Maluti Mountains first traced by Orpen ( 1874) and then by Vinnicombe ( 1976).
This is obviously not a "realistic" painting. It is a painting of a shamanistic visionary hunt, the explanation of which was provided by Orpen's San informant, who said that shamans used buchu to attract the rain bull. In Yates et al. (1990, fig. 25) the "bull" in the foreground
Fig. l. A painting of hunters hy Stow ~md Bleck (19.")).
is clearly responding to the scent (look at the nol>tril~) of the "buchu" plant? soaked skin? held in front of him.
This raises a question. Would shamans employ hunter tactics in visions whieh did not have ~omc:: sort of precedent in actuality?
Two points are to be noted. The animal is a rainbull, not an eland. Second, its posture l>uggt:.\ls that it is moving forward, attracted hy the scent of somdh ing offered by a kneding unaggrel!sive figure. This figure seems to be wearing a cloak. In the ba~kground we :,ee a successfully captured rainhul.l being led by several figures, the first of which has l:tn odd headdress and a winged cloak. Smell or scent cer1ainly seems important in this painting of a vision. Neither of the rainbulls i~
bleeding from the mouth or the nose.
This usc of buchu or other scented suhstancc to attract visionary rainhulls forms a revealing contrast with several paintings in which a hunter/shaman extends a wand, flywhisk or merely his hand to the nostrils/mouth ofwhat is quite unmistakably an dand, who is stationary, or retreating, bleeding or emitting mucous from the mouth/nose region.
Figure 2 is a detail of a painting reproduced by Vinnicombe (1976) in which an eland, unwounded, but bleeding from the nose and not under attack, il>
approached from the front by a man with bow and arrow and a curious long staff.
The possibility that the wand-like object contains Hertia pallens or a similar poisonous hc::rh to encourage the bleeding must be considered. It has heen suggested that the wand-like ohject may he a tly whisk, however it is probable that in this picture it is not whisking tlies away, but inducing nose-bleed.
The same action is presented in Yinni<:omhe 1976:349 fig. 248. In this case the figure is clearly in a distinguishing costume, possibly of a therianthropic k'ind as witnessed by his headgear of horns. He:: is a shaman.
There is no obvious blood, but ::~ suggestion of mucous.
It is possible that this is a shaman's trance vision.
The following alternative scenario is proposed for
'·
Fig. 2. An eland, apparently hleedin~ from the nose, with a hunter cxlcndin,.: a rigid wand-like ohject towards its head. (Vinnicumhe 1976:346, lig. 247).
consideration.
The eland is a huge animal and easily run to a
~tandsti II, hdplesl> wh~n exhausted, unahl·e to use its
horn~. There might have heen special merit in killing an dand hy running it down, without spear or arrow. The only disadvantage of this method i~ that thc:re might be no sacred hlood available to ~ignal the:: death and to ~mc::ar
on the hunter's or hunters' facc{s) to symbolise th~ir
participation in that dc::ath. To guarantee that there is blood avail<~ble, and to encourage the nose-bleed, the chief hunter/shaman approaches the exhausted beast from the front and place!> some toxic soaked absorbent material in front of its nose anJ mouth. One advantage of this, from the shaman's point of view, il> that it would demonstrate hi~ power~ in a ~pectacular fashion.
Th~ technique could he u~ed in addition to the normal weapons of bow and arrow, as clc::mon~trated in the Figure 3. This could depict the tina! stage::!> of an eland hunt. One unarmed hunter is kneeling HI 1he bleeding head of the dand, possihly applying the "snuff" to encourage the bleeding, possihly to collect blood for smearing purposes. His other hand is holding the eland between the horns. Another seems to be stabbing the eland. Blood seems to be dripping from its neck and chest. A third is climbing onto its back. These three hunter figures are <11l wearing cloaks. The two figures in the background are cloakles!>, do not seem to be armed, and may be bleeding from the no~e. The face of one may be darkened by blood smear, but may not be human at all. Their abnormal po~ture!> suggest the:: abnormal world of trance dance. They may not, in fact, be part of this picture. They are facing the wrong way, and are dressed differently.
Perhaps the best evidence for the use of snuff to cause or encourage nose-bleed is from Camp Siding, Tylden District, Eastern Cape (also reproduced in Townley- Johnson 1979:39, tig. 35) (Fig. 4).
The dand is on its (al,t legs, staggering. The hunter is so confident that a beast in this state is heirless that he approaches it naked and unarmed. His hand is extended
0 30
-=-
mmFig. 3. A copy of a p:~intin~.: from the N:~tal Draken!>herg.
(Redrawn from Pa~.:er, after Yates eta/. 1990:44, fi~ 21).
Fig. 4. A pamtm~.: of a man and an eland from Camp Siding, Tylden District, Eastern Cape.
directly to the nose:: from which hlood is heing "sneezed".
This view challenges previously hdd interpretations of the picture. The Townley-Johnson ( 1979:39) caption rejects the literal views that the man is feeding a domes- ticated eland, or is mistakenly offering arrows instead of vegetation, and veers towards a shamanistic approach.
"In sevc::ral paintings of human tigures with ·rain animals', the human hold~ out to the animal 'l>
muzzle objects that look like tufts of leaves.
These have:: heen interpreted as the hu~.:hu or aromatic llerhs that the Bu~hmen helieved would calm the rain animal" (Townley-Johnson 1979:
40).
This may he true enough, hut does not apply here.
Rain animals are sddom depicted as such realistic e'lands, and whatever is being ofl"ered does not look like tufts of leaves. The offering has not heen accepted hy the:: animal.
This may he a picture of a hunter/shaman inserting poisoned snuff, causing nose-bleed. The animal i~
without doubt an eland, staggering on its last legs, not a rain animal.
But the most convincing evidence-is from a painting from the Kamhc::rg (Fig. 5). In the off-mid-foreground is an eland with a defeated po~ture and a white drooping head which is, it seems, hc::ing held hy one hand of a
Fi~. S. A painting of eland and figures from Kamherg.
hunt~r. leaving the other free to insert the snuff. This active figure is wearing a cloak.
Then, ahove this ~cene, and to the left, is the most interesting of all nose-bleed-inducing depictions. A fine dand, staggering backwards, is faced by a group of four figures. Thc::y are all dancing (or kneeling?). The first is holding a wand firmly in contact with the animal's nose.
The next figure holds an unidentified object in one hand (a container for the expected flow of blood?). The remaining two are hand-clapping and no doubt singing.
CONCLUSIONS
It is suggested that all thc::se pictures are of shamanistic
hunte~ at work, either painted by themselves or artists working to their instructions. The incidents depicted are all of the employment of a technique to induce a nose- bleed in an eland to secure its sacred blood. No other animal i::; depicted as being treated in this familiar way.
The conclusion is that this is a religious hunt of the animal most sacred to the People of the Eland.
REFERENCES
Arbousset, T. & Daumas, F. 1846. Narrative of an exploratory tour to the north-east of the Colony of the Cape of Good Hope. Translated from the French of the Revd T. Arbousset hy John Croumbie Brown.
Cape Town: Robertson.
Biesele, M. 1993. Women like meat. Johannesburg:
Witwatersrand University Press.
Bleek, D.F. 1928. The Nharon: A Bushman tribe of the central Kalahari. Camhridge: Cambridge University Press.
Bleek, D.F. J 935. Beliefs and customs of the /Xam Bushmc::n, Part VII, Sorcerers. Bantu Studies 9: 1-47.
Bleek, D.F. 1936. Beliefs and customs of the /Xam Bushmen, Part VIII, More about sorcerers. Bantu Studies 10: 1 63-199.
Broster, J.A. & Bourn, H.C. 1981. Amagqirha: religion, magic and medicine in the Transkei. Goodwood: Via Afrika Ltd.
Drax, A., Knuttgen, H.G. & Titte, L.K. 1988. The Olympic book of sport medicine. Oxford: Blackwell.
Jolly, P. 1996. Symbolic interaction between Black farmers and South-Eastern San. Current Anthropology 37(2): 277-306.
Katz, R. 1982. Boiling energy: community healing among the Kalahari !Kung. Cambridge, Mass.:
Harvard.
Lewis-Williams, J.D. 1981. Believing and seeing.
London: Academic Press.
Lewis-Williams, J.D. 1988. The world of man and the world of spirit: an interpretation of the Linton Rock Paintings. Margaret Shaw Lecture 2. Cape Town:
South African Museum.
Lewis-Williams, J.D. 1992. Ethnographic evidence relating to 'trancing' and 'shamans' among northern Bushmen. South African Archaeological Bulletin 47:56-60.
Lewis-Williams, J.D. & Dowson, T.A. 1994. Aspects of rock art research: a critical retrospective. In Dowson,
T.A. & Lewis-Williams, J.D. (eds). Contested Images:
diversity in Southern African rock art research. Johannesburg: Witwatersrand University Press, pp. 201- 211.
Orpen, J.M. 1874. A glimpse into the mythology of the Maluti Bushmen. Cape Monthly Magazine (n.s) 9(49): 1-13.
Smith, C.A. 1966. Common names of South African plants. Botanical Survey Memoir 35.
Stow, G.W. & Bleek, D.F. 1930. Rock paintings 10
South Africa. London: Methuen.
Townley Johnson, R. 1979. Major rock pamtmgs of southern Africa. Cape Town: David Philip.
Torg, J.S., Welsh, R.P. & Shepherd, R.J. 1990. Current theory in sports medicine. Ontario: B.C. Decker.
Vinnicombe, P. 1976. People of the eland: rock paintings of the Drakensberg Bushmen as a reflection of their life and thought. Pietermaritzhurg: University of Natal Press.
Watt, J.M. & Breyer-Brandwijk, N.G. 1962. Medicinal and poisonous plants of southern and eastern Africa.
Edinburgh: Churchill Livingstone.
Yates, R., Manhire, A. & Parkington, J. 1990. Pictures from the past: a history of interpretation of rock paintings and engravings of Southern Africa.
Pietermaritzhurg: Centaur.
